Machine-translated by Google. Questions are authored in English.
Top

Discussion

A constraint solver takes seconds on most instances of a problem but occasionally runs for hours.

What is the most appropriate response?

  • A.Reduce the number of variables in the model
  • B.Replace the systematic search with local search
  • C.Introduce randomised restarts with learning
  • D.Increase the strength of the propagation used

Answer: C

That pattern is the signature of a heavy tail caused by unlucky early branching, and restarting with different random choices while keeping the learned no-goods truncates it directly. Stronger propagation helps generally but does not address the tail, model size is fixed by the problem, and local search would lose the ability to prove unsatisfiability. Matching the remedy to the shape of the runtime distribution, rather than to its average, is the diagnostic skill here.

No comment is present. Be the first to comment.
Loading…

Post your comment

Please log in or register to post a comment.